Grammar schools should be judged by a tougher benchmark than other state schools, with greater scrutiny of how many pupils achieve the very top GCSE grades, according to a report from the largest network of schools in the country.
Academically selective schools should be held to account on the proportion of pupils that achieve five A*-A grades including English and maths, rather than the five A*-C measure currently used by the government, it suggested.
The recommendation is made in a study by the Schools Network, formerly the Specialist Schools and Academies Trust, which claims that the existing benchmark fails to challenge grammar schools or properly reflect their intake of pupils.
Education secretary Michael Gove has previously stated his desire for various groups, including teaching unions, to introduce their own performance measures that could be used in school league tables.
The author of the report, Professor David Jesson of the University of York, who is an associate director of the Schools Network, said there was considerable regional variance when it came to grammar schools' GCSE results.
"Grammar schools should expect to achieve high levels of performance for their pupils, and most do. There are, however, substantial differences between grammar schools' outcomes, which tend to go unnoticed in the standard performance tables," Professor Jesson said. "If we are genuinely committed to the idea of excellence for all, we need a new way of measuring the performance of these schools and making sure that every pupil reaches their full potential."
Professor Jesson said that, while 58 per cent of pupils across the country currently achieve the government's five A*-C benchmark, 55 per cent of grammar school pupils hit the five A*-A measure. Despite this superior performance, supporters of the grammar school movement have described the recommendation of a tougher benchmark as "grossly unfair".
Nick Seaton, an executive member of the National Grammar Schools Association, a lobby group that campaigns for the expansion of selective state schools, said that all schools should be judged by the same measure.
"If you're going to have a standard, then all schools should be measured by the same standard," Mr Seaton said. "People would expect grammar schools to produce excellent results, but having said that, it's ridiculous to suggest they should all achieve A*s and As.
"It's not grammar schools that need looking at, it's the comprehensives and academies whose results are atrocious."
The 19 grammar schools in outer London saw the highest proportion of pupils gain at least five A*-A grades, with 75 per cent achieving the measure in 2011. The lowest proportion was in the rest of the South East, with just 40 per cent of pupils managing to hit the target, although this was partly due to the region having more grammar schools.
Brian Lightman, general secretary of the Association of School and College Leaders, said that the Schools Network was right to point out the difference between grammar and comprehensive schools, but called for better value-added measures rather than a separate floor target. "The Schools Network has a point," Mr Lightman said. "You would expect a much higher level of performance to reflect the selective nature of the schools.
"But I wouldn't want to see a separate set of measures as it would just add to the confusion. Expectations in any school should not be subjected to raw data, but on progress and the value added - that would address the issue."
58% - Proportion of state-school pupils gaining five A*-C GCSEs including English and maths in 2011
55% - Proportion of grammar school pupils gaining five A*-A GCSEs including English and maths
164 - Number of grammar schools in England.